PERBANDINGAN ALGORITMA C4.5 DAN NEURAL NETWORK UNTUK MEMPREDIKSI HASIL PEMILU LEGISLATIF DKI JAKARTA

General elections are a means of implementation of the sovereignty of the people in the Unitary State of Indonesia based on Pancasila and 1945 Constitution. Elections held in Indonesia is to choose the leadership of both the president and vice president, member of parliament, parliament, and the DPD. In this study comparison of data mining methods, namely C4.5 and neural network algorithm is applied to both the data legislative candidates to be elected to the legislature and were not selected. C4.5 algorithm is one of the algorithms in a decision tree method that converts the data into a decision tree using the entropy calculation formula. While the neural network algorithm is a method like human neurons to find the best path. From the test results to measure the performance of both methods using cross-validation test method, confusion matrix and ROC curves is known that the neural network has the highest accuracy value which is equal to 98.50%, followed by the C4.5 algorithm method with 97.84% accuracy values. AUC values for the neural network method showed the highest value of 0.982 and a decision tree algorithm with a value of 0.970